Is there a vacuum pumping technology availible that can...

Discuss the technical details of an "open source" community-driven design of a polywell reactor.

Moderators: tonybarry, MSimon

chrismb
Posts: 3161
Joined: Sat Dec 13, 2008 6:00 pm

Post by chrismb »

icarus wrote:Solution: After however many days, hours operation it takes to contaminate reaction
at 500MW? milliseconds.. then how long to pump down again?

D Tibbets
Posts: 2775
Joined: Thu Jun 26, 2008 6:52 am

Post by D Tibbets »

Keep in mind that this is based on limited knowledge, but an approach might to keep the fusion products ionized. Have every thing possible outside the magrid carry a mild positive charge, including the vacuum vessel wall (except the decelleration grid). Decellerate the ions to harvast most, but not all of the kinetic energy, then with more periferal positive grids direct the ions into ports with cone shaped baffles (need to be charged?)to prevent significant backstreaming. Have a second vacuum tank (ballest) that the ports empty into, and then allow the ions to neutralize on the grounded walls. I'm hoping that these segragated neutrals would have a pressure significantly higher than the ions in the reaction vessel, greatly enhancing the pumping efficiency of the large pumps that drain this secondary volume.

Since the fusion ions (weather from DD or PB11 fusion) remain charged in the reaction vacuum vessel they could not reenter the core due to the large positive potential on the magrid (and the magnetic fields). This would prevent the reaction space from being contaminated by the charged fusion ash and the limiting facter would be arcing from all the external ions. Obvoisly the 'ion pumps' would have to be highy efficient. In effect you would have a three stage pumping system instead of a two stage pumping system - 'ion pump'- diffusion or turbomolecular pump- mechanical pump.

How are Tokamaks suposed to handle this problem?



Dan Tibbets
To error is human... and I'm very human.

chrismb
Posts: 3161
Joined: Sat Dec 13, 2008 6:00 pm

Post by chrismb »

D Tibbets wrote:Since the fusion ions (weather from DD or PB11 fusion) remain charged in the reaction vacuum vessel they could not reenter the core due to the large positive potential on the magrid
Those that stay as ions, sure, that might be true. D'you think ALL ash ions will not neutralise? It is a population thing, and with 1E21 alphas pouring out of the reaction volume per second, so if milli-fractional amounts of that population neutralise and find their way back into the reaction volume, they will kill it cold.

The next problem is that as those neutrals enter the reaction volume, where, in the potential field, are they ionised? Unless they are ionised either right at the edge, or right at the centre, then they're going to deflect fuel ions tangentially, thus causing rotation/orbiting, and you can't get uniform currents orbiting spherically with a uniform distribution of "theta" AND "phi". That'd be a 'hairy ball' type problem. SO it'd end up chaotic, leading to thermalisation.
D Tibbets wrote: How are Tokamaks suposed to handle this problem?
Not sure what you mean. There are no problems to this, so long as you can deal with a thermalised end result. A tokamak starts off trying to *get* a thermalised result.

As for ash extraction, for fortuitous reasons not well understood, the 4He tends to selectively aggregate around the divertors, where the vacuum pumps are and where they get cooled down by dumping their thermal load into those divertors. It's a win-win for tokamak on that issue, though as this 4He accumulation is not well understood, I guess there is no guarantee it will extend to ITER. That'd be a shock [show-stopper] for them if they found out that 4He collected elsewhere in the plasma!

KitemanSA
Posts: 6114
Joined: Sun Sep 28, 2008 3:05 pm
Location: OlyPen WA

Post by KitemanSA »

chrismb wrote: The mechanism cannot afford to ionise helium ash that wanders back into the well, but it will do this exactly if the pumps miss ANY of the ash.
Sorry for my osmiumnity, but why? The energy to ionize is small, the ash circulates and nudges things this way and that, and eventually does get upscattered out of the well (hopefully at a low rate). Why is it bad to have the "ash" there?

chrismb
Posts: 3161
Joined: Sat Dec 13, 2008 6:00 pm

Post by chrismb »

KitemanSA wrote:Sorry for my osmiumnity
I know my fair share of big words - but what is 'osmiumnity'?
KitemanSA wrote:The energy to ionize is small
Yes, but where does the ionisation take place? If you are hoping to maintain non-maxwellian distributions, you cannot chance having ions being created anywhere other than at the top of the potential well. Otherwise, you'll get thermalisation, and we're then talking a plain-old game of magnetic confinement.

I'm not sure what argument might better satisfy you that this is a wholly unacceptable situation. Perhaps I could also point to the electrons - once you've biffed the electrons off these, then, ions, somewhere uncontrolled within the well, what do those high enery electrons do then? One thing for sure, they're going to get bounced around alot before finding some magnetic minimum and will be emitting powerful brems (at the well potentials we're talking about for p11B) until they cool down. Is that argument more convincing?

KitemanSA
Posts: 6114
Joined: Sun Sep 28, 2008 3:05 pm
Location: OlyPen WA

Post by KitemanSA »

chrismb wrote:
KitemanSA wrote:Sorry for my osmiumnity
I know my fair share of big words - but what is 'osmiumnity'?
The densest known element is either osmium, or between my ears. :D
chrismb wrote:
KitemanSA wrote:The energy to ionize is small
Yes, but where does the ionisation take place? If you are hoping to maintain non-maxwellian distributions, you cannot chance having ions being created anywhere other than at the top of the potential well. Otherwise, you'll get thermalisation, and we're then talking a plain-old game of magnetic confinement.
Actually, near the top of the well but inside the MaGrid is exactly where I expect it to happen. That is where the neutrals (not expecially zippy so they "drift" around there) meet the large flux of electrons being reflected by the MaGrid (maximum zippy) and go bang! No?

chrismb
Posts: 3161
Joined: Sat Dec 13, 2008 6:00 pm

Post by chrismb »

KitemanSA wrote: The densest known element is either osmium, or between my ears. :D
Ah! I see. But no need for such self-critique, the question is fair - and there will be some tolerance permitted for a neutral population, but what it is, is the question.
KitemanSA wrote:
chrismb wrote: where does the ionisation take place?
Actually, near the top of the well but inside the MaGrid is exactly where I expect it to happen. ... No?
I'm certainly in no position to say yes or no, but the collisions are meant to happen mostly in the centre of the reaction voulme, and the free electrons constrained further to the interior (I thought). Why would ionisation happen more where the ion density is less, and further away from the region of slightly-over-net electrons?

If there were electrons hanging around the edge at an energy level that would induce ionisation, surely then they'd also be around the right energy to cause neutralisation with those slow, direction reversing ions also? the cross-section for neutralisation and ionisation are much the same (~10E-16cm2) for these lowish energy levels. Again, what about the electrons from the neutrals? They'd peel off at bang-on the energy to best neutralise some of those slow, annealing ions, wouldn't they?

I don't know and cannot state an answer, but to imagine this reciprocating flow of fuel ions will always cause an ionisation of a neutral right at its least dense edge, without any consequent radiative-lossy recombinations, hmm.... is the future of grid-supplying polywells really relying on this apparent very weak assumption?

KitemanSA
Posts: 6114
Joined: Sun Sep 28, 2008 3:05 pm
Location: OlyPen WA

Post by KitemanSA »

From a wide variety of sources, none that I can quote right now, I have gotten the impression that it is the electron beam that ionizes the neutrals, fuel and ash. That should happen where the electrons are plentiful and fast, the edge of the well. No?

MSimon
Posts: 14332
Joined: Mon Jul 16, 2007 7:37 pm
Location: Rockford, Illinois
Contact:

Post by MSimon »

The top of the well (near the grids) is where the ion density will be highest. The magnetic fields insure that.
Engineering is the art of making what you want from what you can get at a profit.

ohiovr
Posts: 409
Joined: Mon Mar 02, 2009 6:36 pm
Contact:

Post by ohiovr »

I have found a turbo molecular pump that can pump 1,800 liters per second of helium (in an already high vacuum). We could put one of these things on each face of a geodesic sphere. But there are 2 problems that I can think of already:
1) where do you put the pump that gets it out of the way of the equipment destroying neutron flux
2) what ever helium atoms don't go directly towards the pump intake get bounced back into the reaction chamber with a high likely hood of straying into to reactor's center poisoning the reactions.

I thought of one way to improve but maybe not solving the problem of #2. The pump intake could be surrounded by a positively charged wire mesh which repels ionized ash towards a center where the pump intake would be. I don't have any confidence this would work as I am not an electronic engineer.

The solution to #1 I am afraid will be supremely complicated if even possible. Perhaps it would be possible for a duct to make a 90 degree angle back to the turbo molecular pump where the face is neutron shielded with a moderator and boron 10.

ohiovr
Posts: 409
Joined: Mon Mar 02, 2009 6:36 pm
Contact:

Post by ohiovr »

D Tibbets wrote:Since the fusion ions (weather from DD or PB11 fusion) remain charged in the reaction vacuum vessel they could not reenter the core due to the large positive potential on the magrid (and the magnetic fields).
Would the magnetic field also repel the helium ions? The alpha particles, immediately after they are formed (according to rnebel), make about 1000 trips around the magrid coil before exiting. That's with energies of a few MeV. Wouldn't the low energy (1< eV) helium ions get trapped by the magrid field and begin to pile up, circling the magrid infinitely? Or would the several 15-500K volt charge on the magrid be sufficient to toss away all the low energy ions?

Of course theres always the neutrals to deal with. But like you say the shell could be positively charged.

And could it be possible to de-energize the magrids periodically to flush them of what ever ash piles up there? Too great an energy loss perhaps?

ohiovr
Posts: 409
Joined: Mon Mar 02, 2009 6:36 pm
Contact:

Post by ohiovr »

maybe we could use something like this:

http://en.wikipedia.org/wiki/Ion_pump

KitemanSA
Posts: 6114
Joined: Sun Sep 28, 2008 3:05 pm
Location: OlyPen WA

Post by KitemanSA »

With properly selected venetian blind slats, the ion pump would automatically be part of the system.

Personally, I think that having a graphene membrane between the slat and the plasma would allow reasonable capture of those ions that don't embed themselves.

kurt9
Posts: 566
Joined: Mon Oct 15, 2007 4:14 pm
Location: Portland, Oregon, USA

Post by kurt9 »

Cryo-pumps are often used for UHV systems such as MBE. You can also add a cryo-trap to either a turbo or diffusion pump to increase the pump rate.

93143
Posts: 1131
Joined: Fri Oct 19, 2007 7:51 pm

Post by 93143 »

chrismb wrote:I'm certainly in no position to say yes or no, but the collisions are meant to happen mostly in the centre of the reaction voulme, and the free electrons constrained further to the interior (I thought). Why would ionisation happen more where the ion density is less, and further away from the region of slightly-over-net electrons?
It sounds like you've misunderstood the proposed geometry of the wiffleball. The idea of ions being accelerated towards a small central magnetically-confined electron-rich region, which they are initially outside, is wrong. That's what it sounds like you are describing.

The wiffleball is a large spherical region containing a ball of slightly electron-rich plasma, which due to electron currents excludes the magnetic field. The electrons are at high energy at the edge and mostly radially directed. This allows them to penetrate to the interior of the wiffleball rather than forming a sheath, and leads to the formation of a deep potential well.

The ions are FORMED at that very same edge, possibly via ECR tuned for the B-field halfway down the edge gradient, or some such. They are thus low-energy and thermal at the edge. Since the bulk of the wiffleball below them is electron-rich (the aforementioned deep potential well), they tend to fall inwards.

Both electrons and ions would be contained by the strong B-field in a power reactor. Basically, as they enter the non-zero B-field region, their gyroradius goes down dramatically and they end up pulling a U-turn and heading back into the wiffleball (the gyroradius variation is symmetric along both halves of the U-turn, so they don't get trapped on a field line and leave unless something bumps them). For ions, this mostly applies in the case of significant upscatter, since normally the ions are electrostatically confined by the E-field generated by the excess electron population. Anyway, the upshot is that the ball of electrons and the ball of ions are the same size, and occupy the same volume.

Since there is a slight excess of electrons over most of this volume, the electrons slow down as they approach the centre. However, the ions speed up as they approach the centre (obviously), and the result is that there is a high concentration of high-energy ions in the central core region of the wiffleball (even though an individual ion spends only a small fraction of its reactor lifetime in this region). This creates a net excess of ions in the core, which causes electrons to stop slowing down and speed up again as they get very close to the centre and results in some bremsstrahlung. This is the 'virtual anode' effect.

The ions spend the majority of their time near the edge of the wiffleball, due to their low velocity there, which results in a high ion density locally. Because of the low relative ion velocities as compared with the core or the two-stream acceleration/deceleration region (the "mantle"?), the edge region is also where the ion-ion collisional cross-section is the highest, so for the ions, thermalization in the edge region dominates thermalization elsewhere. This is what we call the "annealing" effect.

That's the static picture, as I understand it. POPS will modify this somewhat. Partial relaxation due to collisions should only blur it.

Apparently with ECR you can get very efficient ionization, such that neutral ash making it a significant distance into the wiffleball would be very rare. Also, the electrons are mostly fast and quite populous in the edge region, so neutrals should have a pretty short lifetime whether ECR is used or not...

Post Reply